Vitamin D insufficiency is a common health problem in elderly individuals, who also have a high prevalence of neurodegenerative diseases. Vitamin D is primarily produced in the skin on exposure to UV-B radiation and is found in limited food sources1,19
; advancing age, obesity, avoidance of sun exposure, residence in northerly latitudes, and darker skin pigmentation are associated with increased risk of vitamin D deficiency. Patients with chronic neurodegenerative diseases frequently have many risk factors for vitamin D insufficiency. Consistent with our a priori hypothesis, we found the prevalence of vitamin D insufficiency in the PD cohort to be significantly higher than in the control group. Surprisingly, the prevalence of vitamin D insufficiency in the PD cohort was higher than in the comparison cohort with AD.
The prevalence of vitamin D deficiency (25[OH]D, <20 ng/mL) in our PD cohort (23%) was lower than that previously observed in a Japanese PD cohort13
(50 of 71 patients with PD [85%]). Both insufficiency (36%) and deficiency (10%) in our control cohort were also less prevalent than reported in controls from a study of prostate cancer risk20
(58% insufficient [<32 mg/mL] and 16% deficient [<20 ng/mL]). Because all our study participants resided in southern latitudes and were predominantly white and a greater proportion of our samples were collected in the summer to fall, a lower prevalence of hypovitaminosis D in all our cohorts is not unexpected. Also, the Japanese participants were drawn from a hospital population, whereas participants in this study were recruited from an outpatient clinic. Previous studies14,21
indicate that hospitalized patients have lower mean vitamin D levels than age-matched participants recruited from the community. The higher prevalence of insufficiency in the PD cohorts compared with the control cohort is not unexpected given that PD may cause patients to have decreased activity levels and lower sunshine exposure.
However, the lower vitamin D levels in the PD vs AD cohort are intriguing. The typical course of AD is shorter than that of PD,22,23
and PD patients experience mobility problems more frequently than AD patients. Both factors could make a PD patient less likely to get sun exposure and account for the higher prevalence of vitamin D insufficiency. Supporting this notion, more severely affected PD patients with a longer mean (SD) disease duration (Hoehn and Yahr stages III–V, 7.1[3.8] years) reportedly had a higher prevalence of 25(OH)D deficiency than less severely affected patients with shorter disease duration (Hoehn and Yahr stages I–II, 4.1[2.3] years).13
In contrast, although the range and mean of symptom duration in our PD and AD cohorts were consistent with what one would expect to find in a neurology clinic, 25(OH)D insufficiency did not correlate with symptom duration in either cohort (AD or PD), suggesting that vitamin D insufficiency may be unique to PD. Alternatively, the possible correlation may be too weak to be detected with our current sample size.
Strengths of this analysis are that the sample size was relatively large (97–100 participants in each cohort) and that the cohorts were matched for sex (which, depending on cultural and social differences, could affect duration of sun exposure and degree of skin coverage or protection when in sunlight), age (known to affect susceptibility to vitamin D deficiency), and APOE
genotyping (which has been postulated to be protective for vitamin D deficiency24
). Age was not limited in our inclusion criteria, thus increasing the generalizability to PD and AD populations with younger and older patients. In addition, diagnoses were made by experienced subspecialty neurologists according to well-characterized criteria. Although this was a retrospective study, participants were identified and data were analyzed as if in a prospective cross-sectional study.
This analysis did not control for vitamin D intake, although in other populations vitamin D intake contributes little to 25(OH)D levels.25
In this cross-sectional study that lacks longitudinal data, we cannot address etiology for the clinical associations discussed. In addition, we do not have uniformly collected measures of PD severity (Hoehn and Yahr stage, Unified Parkinson Disease Rating Scale scores) or anthropomorphic and sun exposure data; therefore, we cannot assess what effect, if any, differences in body mass index and exposure to sunshine might have on these findings. Another weakness is that the portion of plasma samples drawn in the winter to spring and fall to summer were not matched across cohorts. Despite a significantly higher proportion of the samples from the PD cohort being drawn in the summer to fall, we still found higher rates of vitamin D insufficiency in the PD cohort compared with the AD and healthy control cohorts.
In summary, we found that PD patients have a higher prevalence of vitamin D insufficiency compared with patients with AD and healthy controls. These findings support the previously suggested need12
for further studies to assess what contribution a low 25(OH)D concentration adds to the risk of developing PD (vs other neurodegenerative disorders) and to determine whether correction of vitamin D insufficiency and deficiency will improve motor or nonmotor symptoms in PD. Finally, the finding of a high incidence of vitamin D deficiency in the PD and other cohorts highlights the importance of routinely checking the level of 25-(OH)D, particularly in elderly patients, since deficiency is strongly correlated with a higher incidence of osteoporosis, falls, and hip fractures and has been associated with a higher incidence of several forms of cancer and autoimmune disorders.